On the Horizon

For all that GPT4 and ChatGPT have kicked off a round of doomsaying that artificial intelligence will decimate the job market and potentially even result in the extinction of humanity, it's important to remember that thus far, what we have a very powerful and sophisticated version of Autocomplete. General Artificial Intelligence, which is what so many people are worried about, it still some ways in the future; which, to be sure, it's always been.

For right now, what people consider to be AI is of limited use. Attaching a consumer-grade Large Language Model to the Bing search engine results in something that couches a non-answer to the questions I've asked it in natural-seeming language. Which is interesting, but perhaps more proof that it really isn't intelligent, in the sense of having the ability to use the information at its disposal to formulate useful answers to novel questions.

Likewise, a blindingly obvious phishing e-mail landed in my Outlook inbox. One would think that a genuine AI would be immensely helpful in detecting these sorts of things, this providing a market advantage.

So I think that I'll take a pass on the LLM hype for the time being, and see what the next big innovation has to offer. I appreciate the concerns of the people who tell us that we should be preparing now, mainly because at the current state of the art, there is still time to be prepared.